Wan2.6-Image

Model IntroductionImage

Wan2.6-Image: High-quality image generation, strong controllability

Wan2.6-Image is part of the Wan lineup from Qwen, designed for production workloads that need reliable quality at scale. High-quality image generation, strong controllability.

Generate image from Text, Image prompts with fine-grained style and layout control.

Benchmarks

Benchmark (Area)Score*
GenEval (Composition)74.0% accuracy (Eval)
T2I-CompBench (Prompt Adherence)0.70 score (Eval)
PickScore (Aesthetics)18.8 (Eval)
Human Preference (Quality)73.3% preferred (Eval)
Multi-Object (Control)69.6% success (Eval)

API request example

Endpoint /v1/images/wan/generations. Replace <API_KEY> with a key from the console.

Use `/v1/images/wan/generations`. `size`: `1K` / `2K`. Optional reference `images`. Allowed size: 1K / 2K.

curl https://api.ravenrouter.ai/v1/images/wan/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <API_KEY>" \
  -d '{
  "model": "Wan2.6-Image",
  "prompt": "tomato egg stir fry, food photography",
  "images": [
    "https://cdn.wanx.aliyuncs.com/tmp/pressure/umbrella1.png"
  ],
  "size": "1K",
  "n": 1
}'

More models: API Reference → Examples by model

Wan2.6-Image API — $0.02/image | RavenRouter AI